Tiny_Faces_in_Tensorflow icon indicating copy to clipboard operation
Tiny_Faces_in_Tensorflow copied to clipboard

Published 20 hours ago • verified publisher icon cydonia999

Question about runtime

Open srireddy821 opened this issue 7 years ago • 4 comments

What runtime did you observe during your experiments?

srireddy821 avatar Dec 01 '17 14:12 srireddy821

CPU with 16 GB RAM and 4 cores - 6-7 sec per image

On 980 TI GPU - 0.5-1 sec per image

abhi-kumar avatar Dec 08 '17 15:12 abhi-kumar

hello! how to use the GPU test model?Is it modified in the tiny_faces_model.py file?

pangyanhui923 avatar May 15 '18 07:05 pangyanhui923

Simply use tensorflow-gpu (install using pip). If this doesn't work do let me know.

abhi-kumar avatar May 15 '18 11:05 abhi-kumar

@abhi-kumar I have installed the tensorflow-gpu, I want to know how to call gpu and cuda acceleration in the code?

pangyanhui923 avatar May 16 '18 14:05 pangyanhui923